(of secondary data analysis). Why and how.
09 May 2023
deductive reasoning / hypothesis testing
(K. Popper, 1935; K. R. Popper, 1989)
does not follow this logic:
e.g. taking a big data set and correlating many variables
(K. Popper, 1935; K. R. Popper, 1989)
Does preregistration make any difference at all? In the end, the test is the same!
Kaplan & Irving, 2015
Kaplan & Irving, 2015
design: John Manoogian IIIcategories and descriptions: Buster Bensonimplementation: TilmannR, CC BY-SA 4.0, via Wikimedia Commons
1.
2. Great tool to help us protect ourselves from biased inferences and help others recognize that bias has been tried to be avoided.
BTW: registered reports even more so
and objections to these objections
and one affirmation of an objection
| objection | comment |
|---|---|
| it’s a lot of work | it’s mainly front-loading |
| linking the pre-registration when submitting a paper, uncovers the blinding | there are possibilities to blind the prereg on OSF and aspredicted.org |
| secondary data analysis comes after the data collection?? | yes, but you didn’t see the data yet (provide proof of data access date) |
| yes, because you might already infer relations in data from other known publications |
Icons by Font Awesome CC-BY 4.0
footer